What you'll learn

Apply the Tallinn Manual's scale-and-effects test to AI cyber ops and determine whether Article 2(4) or Article 51 thresholds are crossed.
Work through the Nicaragua and Tadić attribution standards and identify where each breaks down when AI systems are in the targeting chain.
Identify the three points where IHL's distinction, proportionality, and precaution principles structurally fail under AI targeting architectures.
Assess whether an AI system's supply chain generates state responsibility under ILC Articles 4, 5, 8, and 16 and where accountability gaps emerge.
Advise on AI sovereignty claims, digital sovereignty assertions, and the governance gaps transnational AI architecture creates across domains.
Use a five-layer framework — actors, thresholds, conduct, accountability, governance — to analyse any novel AI warfare legal problem independently.

Course content

5 sections • 14 lectures • 1h 9m total length

Welcome to the Course4:23
What will students be able to do at the end of this section?
By the end of this section, students will be able to describe the overall architecture of the course bundle, identify the five principal bodies of international law that AI warfare places under stress, and locate each of the 22 course sections within the master argument that AI-enabled conflict has fractured three foundational legal assumptions: that states are the primary actors, that force is physically legible, and that causation is traceable to a human decision. Students will be able to explain why this course is structured as a cumulative legal argument rather than a survey of topics, and will understand how each section builds on the last. They will also be able to articulate the course's analytical method — examining doctrine as it exists, identifying its operational premises, and testing those premises against the specific characteristics of algorithmic conflict — so that they can apply that method independently to novel problems as the field develops.
Description: Introduces the course's purpose, method, and scope. Explains what this course is — a rigorous legal and strategic analysis of AI warfare — and what it is not: a technology course or a policy prescription. Sets out who the course is for, how it is structured, and what professional and analytical outcomes students can expect. Frames the foundational challenge: international law was designed for a world this course will systematically show it can no longer fully govern.
The Complete Mind Map of AI Conflict5:22
What will students be able to do at the end of this section?
By the end of this section, students will be able to describe the overall architecture of the course bundle, identify the five principal bodies of international law that AI warfare places under stress, and locate each of the 22 course sections within the master argument that AI-enabled conflict has fractured three foundational legal assumptions: that states are the primary actors, that force is physically legible, and that causation is traceable to a human decision. Students will be able to explain why this course is structured as a cumulative legal argument rather than a survey of topics, and will understand how each section builds on the last. They will also be able to articulate the course's analytical method — examining doctrine as it exists, identifying its operational premises, and testing those premises against the specific characteristics of algorithmic conflict — so that they can apply that method independently to novel problems as the field develops.
Description: Presents the five-layer analytical framework that organises the entire bundle: actors, thresholds, conduct, accountability, and governance. Shows how every section of the course lives within one of these layers and how the layers interact. Students leave this lecture with a navigational map of the full curriculum — a reference architecture they can use to orient every subsequent lecture and to analyse novel AI warfare problems independently.
Section 1 Quiz: Foundations — Method, Framework, and the Master Argument
The Five-Layer Legal Map: Analysing an AI Warfare Scenario

Classical Sovereignty vs Digital Sovereignty6:57
What will students be able to do at the end of this section?
By the end of this section, students will be able to explain the classical Westphalian sovereignty doctrine — its territorial, population, and governmental components — and identify precisely where each component fails when applied to AI-mediated power projection. Students will be able to distinguish territorial, digital, and cognitive sovereignty as three distinct dimensions of sovereign authority and assess the specific legal challenges AI creates in each. They will be able to apply the Tallinn Manual's Rules 1–4 on sovereignty to transnational AI systems, evaluate the legal cognisability of "AI sovereignty" as a concept, and assess whether weaker states can meaningfully exercise sovereignty in AI-dominated conflicts. Students will also be able to analyse how private corporations operating AI systems of military relevance challenge the state-centric architecture of international law, and explain why sovereign equality under the UN Charter does not resolve the factual asymmetry created by concentrated AI military capability.
Description: Traces sovereignty from the Peace of Westphalia through Articles 2(1) and 2(7) of the UN Charter. Shows precisely where the territorial anchor of classical doctrine dissolves when applied to AI systems that operate across multiple jurisdictions simultaneously. Introduces digital sovereignty — states' attempts to assert authority over data flows, algorithmic systems, and network infrastructure — and examines the structural tension between that assertion and the inherently transnational architecture of AI technology.
Cognitive Sovereignty and Information Power5:37
Description: Examines the third and most legally underdeveloped dimension of sovereignty: a state's capacity to govern the information environment within which its population forms beliefs and makes political decisions. Analyses whether AI-enabled influence operations that systematically manipulate a state's cognitive environment constitute unlawful intervention under the Nicaragua non-intervention standard. Assesses the legal cognisability of "cognitive sovereignty" and its relationship to the prohibition on coercive interference in internal affairs.
Sovereignty Without Borders6:30
Description: Examines three compounding challenges: the transnational exercise of sovereignty through distributed AI architectures, the role of private corporations in redefining sovereign authority over AI systems of military relevance, and the practical inability of weaker states to exercise meaningful sovereignty in AI-dominated conflicts. Applies the Tallinn Manual's sovereignty rules to the specific problem of AI systems whose components are distributed across multiple jurisdictions and whose harmful outputs cannot be traced to any single sovereign.
Section 2 Quiz: Sovereignty — Classical, Digital, and Cognitive Dimensions
Digital Sovereignty Assessment: Legal Memorandum

Understanding Article 2(4)5:22
What will students be able to do at the end of this section?
By the end of this section, students will be able to apply Article 2(4) of the UN Charter to AI-enabled operations and determine whether a given operation crosses the use-of-force threshold. Students will be able to distinguish a "use of force" from an "armed attack" and explain the legal significance of that distinction for victim state responses. They will be able to apply the Tallinn Manual's scale-and-effects test to cyber and AI operations, assess whether economic or infrastructural disruption can qualify as force, and evaluate whether preemptive cyber disruption constitutes aggression under UN General Assembly Resolution 3314. Students will be able to explain how hybrid warfare tactics — combining cyber, information, and AI operations below any single threshold — systematically exploit the definitional gaps in Article 2(4), and assess whether AI-enabled multi-party conflicts qualify as international armed conflicts for purposes of IHL application.
Description: Examines Article 2(4)'s drafting history, textual scope, and operational limits. Explains why the provision was calibrated for kinetic, state-to-state military operations and identifies the specific definitional gaps that AI-enabled conflict exploits. Introduces UN General Assembly Resolution 3314's Definition of Aggression and assesses its application to contemporary AI-enabled operations. Distinguishes use of force from armed attack and explains why that distinction determines the legal options available to victim states.
When Does a Cyber Operation Become Force?5:15
Description: Applies the Tallinn Manual's scale-and-effects test to cyber and AI operations. Works through the key factors — severity, immediacy, military character, state involvement, reversibility — using concrete operational scenarios including power grid disruptions, financial system attacks, and communications network degradation. Examines the contested zone between economic coercion and use of force, and identifies the strategic incentives states have for keeping the threshold deliberately ambiguous.
Algorithmic Escalation and Autonomous Attacks5:03
Description: Examines the most acute version of the threshold problem: when AI systems themselves generate or accelerate the escalatory dynamics that produce armed conflict. Analyses algorithmic escalation — where interacting AI systems on opposing sides compress decision cycles below human oversight — and assesses its implications for war classification. Evaluates proxy warfare and hybrid operations against the aggression definition and examines whether AI-enabled multi-domain campaigns can be legally classified as international armed conflicts.
Section 3 Quiz: Use of Force — Thresholds, AI Stressors, and Hybrid Warfare
Threshold Analysis: AI Cyber Campaign Assessment

The Right of Self-Defence3:47
What will students be able to do at the end of this section?
By the end of this section, students will be able to identify the conditions that trigger the Article 51 right of self-defence and apply the customary law requirements of necessity and proportionality derived from the Caroline case and confirmed in the Nicaragua and DRC v. Uganda judgments. Students will be able to assess the legal justifiability of anticipatory self-defence when AI systems compress attack timelines below human reaction speeds, evaluate the meaning of "imminence" in AI-enabled threat environments, and analyse the legal implications of pre-delegating defensive strike authority to automated systems. They will be able to assess self-defence claims against non-state actors operating AI systems from third-state territory, determine when a collective self-defence response to AI-enabled attacks risks violating the sovereignty of neutral states, and evaluate whether proportionality doctrine can be applied to defensive responses against non-kinetic AI-enabled harm.
Description: Establishes the Article 51 framework in full: the armed attack trigger, the customary law requirements of necessity and proportionality from the Caroline case, the ICJ's elaboration in Nicaragua and DRC v. Uganda, and the legal conditions for collective self-defence. Examines self-defence against non-state actors and the divided state practice on whether the "unwilling or unable" standard satisfies the armed attack attribution requirement. Sets up the specific AI-related challenges developed in the subsequent two lectures.
AI and Time Compression4:17
Description: Examines how AI-enabled conflict compresses attack timelines below the human decision speeds that anticipatory self-defence doctrine presupposes. Analyses the Caroline "no moment for deliberation" standard against the millisecond timescales of AI-enabled strikes. Assesses the legal implications of pre-delegating defensive strike authority to automated systems, including the proportionality and necessity problems created when contextual human judgment is removed from the authorisation chain. Evaluates whether "continuous self-defence" doctrines can address persistent AI-enabled threat campaigns.
Collective Self-Defence and Third-State Implications4:15
Description: Examines three frontier questions in collective self-defence as applied to AI warfare: when AI system provision to an ally constitutes participation in collective self-defence triggering legal obligations; when collective self-defence responses to AI-enabled attacks violate the sovereignty of neutral states through whose digital infrastructure the response routes; and how proportionality is calibrated when the original attack was non-kinetic, AI-mediated, and difficult to quantify. Assesses alliance obligations when AI system deployments by an ally may themselves be generating IHL violations.
Section 4 Quiz: Self-Defence — Imminence, Proportionality, Collective Defence, a
Self-Defence Memorandum: Advising a Government After an AI False Positive Strike

The Attribution Problem3:41
What will students be able to do at the end of this section?
By the end of this section, students will be able to apply the ILC's 2001 Articles on State Responsibility to AI-mediated state support in armed conflict, distinguishing the Nicaragua effective control standard from the Tadić overall control test and explaining the evidentiary and doctrinal implications of each. Students will be able to identify the specific attribution gaps that AI systems create — including the false flag problem, the private actor problem, and the autonomy gap — and assess how each undermines the state responsibility framework. They will be able to evaluate whether a state's failure to regulate AI systems within its jurisdiction generates international responsibility under due diligence norms, determine when logistical or intelligence AI support to a conflict party rises to co-belligerency, and explain why the structural architecture of AI military contracting is specifically designed to maintain legal distance between state principals and harmful AI outputs.
Description: Introduces the ILC Articles on State Responsibility and their key attribution grounds — Article 4 (state organs), Article 8 (direction and control), Article 11 (acknowledgement and adoption). Presents the Nicaragua effective control test and the Tadić overall control alternative. Applies both standards to AI-mediated state support in conflict — autonomous weapon provision, algorithmic intelligence assistance, remote targeting support — and shows precisely where each standard breaks down when the harmful act is generated autonomously by an AI system.
Deep Attribution Challenges4:34
Description: Examines the three-level gap between technical attribution, political attribution, and legal attribution in AI-enabled conflict — and explains why that gap systematically favours aggressors and undermines deterrence. Analyses the private actor problem: how commercial AI development structures, contractual arrangements, and corporate confidentiality regimes are used to maintain legal distance between states and the harmful outputs of AI systems they effectively control. Introduces the corporate accountability questions developed in full in Course 4.
State Responsibility in Cyber Conflict4:43
Description: Applies the Tallinn Manual's Rules 14–17 on state responsibility to cyber and AI operations. Examines the false flag attribution problem — where sophisticated AI operations are designed to mimic the signatures of other actors — and its implications for wrongful countermeasures against misidentified states. Analyses the co-belligerency question: when does AI system provision, intelligence sharing, or infrastructure support make a third state a party to an armed conflict with full IHL obligations? Evaluates passive tolerance of hostile AI operations under due diligence norms.
Section 5 Quiz: State Responsibility — Attribution Standards, the Corporate Gap,
Attribution and Accountability Analysis: AI Proxy Warfare Scenario

Requirements

No prior legal knowledge required — every doctrine is taught from its primary source
No prior technical AI knowledge required — relevant AI operational characteristics are explained as they arise
A general professional familiarity with either technology, security, law, or policy is helpful but not mandatory
An interest in how international law applies to real, documented AI deployments in active conflict environments

Description

AI systems are already operating in legal environments that international law was never designed to govern. Every engineer who builds a dual-use AI system, every security professional who responds to a state-sponsored cyber incident, and every legal advisor who counsels a defence contractor is operating inside a legal framework they almost certainly have not been taught—one that determines when their work generates state responsibility, crosses an IHL threshold, or triggers criminal liability up the chain of command.

This course closes that gap.

Foundations of International Law in the Age of AI is a rigorous techno-legal analysis of the five bodies of international law that AI-enabled technology is actively stressing in documented conflict environments right now: sovereignty, the prohibition on the use of force, the right to self-defence, state responsibility, and international humanitarian law. Every doctrine is taught from its primary source — treaty text and ICJ judgment — and applied directly to the operational characteristics of AI systems: autonomous targeting, cyber operations, algorithmic influence, and distributed supply chains.

Across 14 lectures organised around a permanent five-layer analytical framework, you will learn to apply the Tallinn Manual's scale-and-effects test to an AI-enabled cyber operation; work through the Nicaragua effective control test and the Tadić overall control standard and identify where each breaks down in AI proxy warfare; assess whether an AI system's supply chain generates state responsibility under ILC Articles 4, 5, 8, and 16; and identify the three structural points at which IHL's distinction, proportionality, and precaution principles fail under AI targeting architectures.

No prior legal knowledge is required. No prior technical AI knowledge is required. What is required is the recognition that these questions are no longer theoretical — they are being decided right now, in active conflict zones, by systems that people in this profession built or advised on.

This is the legal framework your work already operates inside. This course makes it legible.

Who this course is for:

Cybersecurity professionals and threat intelligence analysts who work on state-sponsored cyber attribution, incident response, and AI-enabled attacks — and need the international legal framework that determines when those attacks cross from hostile act to use of force to act of war
AI engineers, ML architects, and software developers building dual-use, defence-adjacent, or government-facing systems who need to understand when their systems generate state responsibility, trigger IHL requirements, or create export control liability
Legal professionals, in-house counsel, and compliance officers advising technology companies, defence contractors, or governments on the international legal obligations attached to AI system development, deployment, and export
Policy analysts, government advisors, and public sector professionals working on national AI strategy, autonomous weapons governance, cybersecurity law, or international security frameworks
Security consultants, GRC professionals, and risk advisors who need to brief boards and executive leadership on the international legal risk exposure of AI systems their organisations operate
International law students, LLM candidates, and legal researchers working at the intersection of technology, armed conflict, and state responsibility

What you'll learn

Explore related topics

Course content

Introduction: The Legal Architecture AI Warfare Is Breaking2 lectures • 10min

Sovereignty: When AI Systems Cross Every Border Simultaneously3 lectures • 19min

Use of Force: When Does an AI Cyber Operation Cross the Legal Line?3 lectures • 16min

Self-Defence: Why AI Eliminates the Deliberation That International Law Requires3 lectures • 12min

State Responsibility: Who Is Legally Accountable When AI Causes Harm?3 lectures • 13min

Requirements

Description

Who this course is for: